Join 3,424 readers in helping fund MetaFilter (Hide)


"An experiment in organic software visualization."
June 18, 2008 8:56 PM   Subscribe

code_swarm, an animated visualization of open source software project commits. e.g.: Python.
posted by signal (18 comments total) 12 users marked this as a favorite

 
neat!
posted by cowbellemoo at 9:11 PM on June 18, 2008


so cool. i love you guido!
posted by Mach5 at 9:19 PM on June 18, 2008


Watched the postgreSQL, which I am interested in using. Neato, and thanks, signal. Bravo to the open source movement!
posted by Turtles all the way down at 10:45 PM on June 18, 2008


Dang it, I was going to post this. I like how the author annotates his videos with notable events in the project histories.
posted by whir at 10:47 PM on June 18, 2008


I have to say, the Python one is surprising, how Guido and a very small group of contributors plug away for years, and then suddenly KABOOM-ITS-THE-CALVACADE-OF-COMMITS.
posted by davejay at 11:36 PM on June 18, 2008 [1 favorite]


Interesting idea, visualizing information. The tool:
"Processing is an open source programming language and environment for people who want to program images, animation, and interactions. It is used by students, artists, designers, researchers, and hobbyists for learning, prototyping, and production. It is created to teach fundamentals of computer programming within a visual context and to serve as a software sketchbook and professional production tool."
Kind of reminds me of the planetary model of the atom (Bohr). Nice post.
posted by sluglicker at 11:47 PM on June 18, 2008


"how Guido and a very small group of contributors plug away for years"

It's called "public source code repository". Before that, we mailed patches to GvR or posted them to mailing lists.
posted by effbot at 12:08 AM on June 19, 2008


(it's the KABOOM-ITS-THE-CALVACADE-OF-COMMITS part that's the public repository, of course. the KABOOM at the end is the migration from SourceForge, when all contributors got new user names).
posted by effbot at 1:35 AM on June 19, 2008


That's fascinating, thanks signal :)
posted by jacalata at 1:46 AM on June 19, 2008


effbot, do you have any insight on why GvR chose to build Python on Windows on a Microsoft compiler instead of a Free one? That one always bugged me.
posted by CautionToTheWind at 4:29 AM on June 19, 2008


This is a really great idea, but high data rates seem to confuse it. Perhaps a somewhat different interface would work better.
posted by DU at 4:39 AM on June 19, 2008


I love watching the time series at the bottom of the screen. A 1/f noise perhaps?
posted by norabarnacl3 at 6:20 AM on June 19, 2008


Dammit, I want to run this against some other projects myself, but the checkout recipe here yields bupkuss. Anyone able to actually get the damn code?
posted by the dief at 6:44 AM on June 19, 2008


Whoo! Go Team Blue!
posted by Eideteker at 8:13 AM on June 19, 2008


Team blue was the documentation team (or guy), which I found fascinating as the "blue" period of Python's development preceded a time of heavy checkins and, as mentioned in the caption, Python's explosion of popularity.

I wasn't into Python back then; are there are Pythoners out there who can corroborate this? Was the time after the massive documentation the time of Python's initial popularity explosion?
posted by sleslie at 9:20 AM on June 19, 2008


Too many things that are awesome all rolled into one project! Advanced data visualizations, the history of python and processing.. Love it.
posted by mike_bling at 1:05 PM on June 19, 2008


the dief - I managed to check out a copy of .. something .. but it doesn't include any repository data files, tools to create repository data files, or even a description of the format (although this is almost verging on being one), so it's not really ready for random people to use yet.

I don't think I really want to look too hard though - the results for the main open source thing I hack on would probably be pretty depressing.
posted by plant at 7:20 PM on June 19, 2008


After some idle hackery while I was cooking dinner, I've come up with a reasonably dumb perl script that wrangles history from a git repo into the xml format this thing uses. The results are, as expected, pretty sad when compared to the bright shiny interesting animations for larger projects with more committers and more interesting code structures. I'm currently running it over the last year of webkit development history, since I happened to have that lying around, and that's looking a bit more interesting.

Here's my script, in case anyone wants to play with it. It generates 'event formatted' files, so you need to replace the call to 'loadRepository' with 'loadRepEvents' in code_swarm.pde.
posted by plant at 7:06 AM on June 20, 2008


« Older Slangin' Liquor in the Hood...  |  Indiana Jones and the Kingdom ... Newer »


This thread has been archived and is closed to new comments